Roy T. Fielding wrote: > ... It may be worthwhile to think in terms of a project's lifecycle:
As a new feature is being invented (i.e., before a component's v1.0...), it is not a bad thing to have multiple teams all working on their own interpretation of the problem and innovating on their proposed solution. In this phase, being different is not a liability as there are no existing expectations to live up to. As these new components start being integrated into systems (and thus get more outside attention...), each distro will need to decide which to use. Firefox -vs- safari -vs- opera -vs- ..., or even solaris-sh -vs- bash -vs- ksh93 are all current examples. Once a distro decides on a specific componant, it is (as it should be) difficult to change behaviors or components. In this phase, multiple incompatible instances of a particular component are seen as a bad thing. I think Roy (et.al) is coming from a "prototyping and initial invention of new stuff" perspective, while Jim (et.al) are more worried about the "once we have made a commitment, how do we keep it" one. -John
